User:SlvrDreams/Sandbox
From Wikipedia, the free encyclopedia
Contents |
[edit] Design Concept
Create a visual mapping of Wikipedia and the blogosphere while using higher-level mathematics to develop dynamic relationships between related topics. As the wiki and blogosphere grows in size, the ability to easily search becomes more difficult. In addition, by visualizing the wiki and blogosphere using dynamic relationships, new innovative connections can be established that may have previously been unknown. I am looking for several programmers that would be interested in starting an open source initiative to create this interface. My background is electrical engineering and I understand the higher-level math; however, my programming skills aren’t so good.
[edit] Related Examples of Visualization
While data visualization is not new, I am proposing that higher-level math be used to create dynamic relationships which will assist in knowledge discovery. Here are several website examples to illustrate the concept:
- We Feel Fine–(Jonathan Harris) A visual mapping of all emotional statements in the blogosphere
- Universe–(Jonathan Harris) A visual mapping of news stories.
- Live Plasma–Relating musicians and movies through visualization
- Wiki Mind Map–New initiative into mapping Wikipedia using the opensource MindMap software.
- CNet
- Discover The Network
[edit] Mathematics to define relationships
Tags and categories can be used to define what topics are related, which statistical analysis (Link Analysis, data mining, and information theory) can be used to define the strength of the relationship.
- Del.icious
- Link Analysis
- Data Mining/Knowledge Discovery
- Social Networking Theory
- Information Theory
[edit] Detailed Math
[edit] Information Theory
A fundamental measure in Information Theory is entropy. Entropy is a measure of uncertainty in a communication system, but it can be used as a weighting value of a random process. Given a node x, calculate an entropy value of each branch and then normalize each value. The resulting value can be used to weight the length/thickness of the relationship.
[edit] Example Images
In order to better visualize the concept, I created several graphic examples. My thought is you could navigate the map by clicking and dragging or by entering a search term. Each node in the map represents a wiki or blog page which opens in a new window.
